Introduction: Standard brush cytology has poor sensitivity for detecting malignancy in biliary strictures. Fluorescence in situ hybridization (FISH) has been reported as a useful adjunctive test in evaluation of biliary strictures. However, the diagnostic performance and definitions of a positive FISH result vary across different studies. This meta-analysis aimed to define the diagnostic performance of FISH in evaluation of biliary strictures.
Methods: A systematic search of Ovid MEDLINE, EMBASE, Cochrane and Scopus databases was conducted by an expert librarian through July 2023. Studies reporting the diagnostic performance of UroVysion®(Abbott Molecular) probe set in adult patients (>18y age) with biliary strictures were included. Sensitivity analysis of studies using (1) only polysomy (>2 copies of ≥2 probes) and (2) polysomy and tetra/trisomy as criteria for positive FISH test were also performed. Bivariate random-effects model was used to summarize the results from individual studies while keeping the two-dimensionality of the data. Receiver operating characteristic curves (ROC) were constructed using a hierarchical summary. Results were reported as sensitivity (Sn), specificity (Sp), positive and negative likelihood ratio (LR). All statistical analyses were performed using the R-software. Results are presented as point estimate (95% confidence interval [CI]).
Results: A total of 17 studies with 1042 FISH specimens were included. Malignancy was identified in 596 (57.2%) cases.. A positive FISH (as defined by the study authors) had a pooled Sn of 56.7% (49.2-63.8%, I²=19.3%) and Sp of 86.5% (76.7-92.5%, I²=19.3%). Negative FISH had a small magnitude of change (negative LR=0.5 [0.4-0.6]) whereas a positive FISH resulted in a moderate magnitude of change in probability of diagnosis (positive LR=4.4 [2.5-7.2]) after testing. Pooled false positive rate was 13.5% (7.5-23.3%, I²=19.3%). Performance of FISH was similar amongst studies with primary sclerosing cholangitis (PSC, n=3). In PSC, pooled FISH Sn was 56.2% (28.3-80.6%, I²=0%) and Sp was 87.5% (46.0-94.5%, I²=0%)

On sensitivity analysis of studies using only FISH polysomy as considered positive (n=9), FISH Sn (50.1%, 44.5-55.7%, I²=18%) was similar however the Sp increased to 94.5% (89.3-97.2%, I²=18%). In those studies that considered tetrasomy/trisomy in addition to polysomy as positive (n=8), FISH Sn increased significantly to 65.1% (57.7%-71.8% I²=0%), however Sp dropped significantly to 75.7% (60.4.-86.4%, I²=0%).

Conclusions: Overall FISH testing for detecting malignancy in biliary strictures has a pooled sensitivity of 55.9% and specificity of 88.1%. These test characteristics are impacted by the definition of a positive test. Reporting of FISH testing requires further standardization to avoid false positive test results.

Pooled diagnostic characteristics of fluorescence in situ hybridization (FISH) for diagnosis of malignancy in biliary strictures stratified by definition of positive FISH test.

Emily Barr Fritcher

Appearances